A crane and hoist gearbox is a low-speed, high-torque unit that must survive repeated starting under full load, frequent short cycles and a load that reverses direction on every trip. The single most common selection error is choosing it from motor power and output speed alone, then discovering that the thermal rating, the brake position and the bearing that carries the load path were never checked. This guide covers the four decisions that actually determine whether a hoist drive survives its first year: duty class, service factor, brake placement and the load path through the housing.

Start With the Duty Class, Not the Motor

Crane mechanisms are classified by the number of operating cycles and the severity of load they see, conventionally written as a mechanism group such as M3 through M8, with M4 and M5 covering most general factory and yard cranes and M7 and M8 covering continuous container handling. The class is not a marketing label: it converts a vague "heavy use" into a number that scales the allowable gear and bearing stress.

Two hoists with identical lifting capacity can require very different units. A maintenance crane that lifts three times a shift sees perhaps ten cycles an hour; a scrap-yard magnet crane may see sixty. Multiply cycles per hour by hours per day, by load spectrum, and the annual revolution count of the output shaft can differ by a factor of fifty. A unit sized on torque alone will pass both, then run its bearings past their rating on the second one.

If the mechanism group is unknown at enquiry stage, state cycles per hour, hours per day, average load as a fraction of rated load, and the expected years of service. Those four numbers let a supplier place the drive in the correct class rather than guess.

Service Factor Is a Multiplier on the Load, Not a Safety Margin

The service factor multiplies the nominal torque to produce the design torque used for gear and bearing rating. It is not a comfort allowance; it is the mechanism that converts the duty class into a stress level the gearing can carry for the required life.

Hoisting differs from travel in how that factor is applied. During lifting, the load is relatively steady and the gearbox sees near-nominal torque for most of the cycle. During lowering, the drive often runs in the opposite direction with the load driving the motor, so the same gearing absorbs torque on the opposite face. Frequent reversing, which every hoist does, means both tooth flanks wear. That is why the service factor for hoist duty generally sits higher than for a comparable conveyor drive of the same power.

For travel drives there is an extra term that catches designers out: the acceleration torque. A bridge or trolley that must reach full speed in a fixed time draws a starting torque well above running torque, and the gearbox sees that peak on every trip. Where a variable frequency drive is used, the starting torque can be controlled and the peak reduced; where a contactor-based starter is used, the peak arrives at every start.

The practical rule: calculate design torque as nominal torque times service factor, then check that the peak starting torque does not exceed the unit's allowable peak rating, typically around twice the nominal rating for a short duration. Passing the average and failing the peak is a common and expensive mistake.

Brake Placement Changes What the Gearbox Sees

Where the brake sits in the drive train decides which components absorb the stopping load.

A brake on the motor shaft is the usual arrangement. It stops the whole train, so the gearbox absorbs the deceleration torque of the load reflected back through the ratio. Because that reflected torque is divided by the ratio, the gearing generally handles it comfortably, but the brake itself must be sized for the full reflected inertia, not just for holding.

A brake mounted on the high-speed input of the gearbox behaves the same way electrically but is easier to service, since it stays accessible when the motor is tucked into a frame.

A brake on the output side, or an independent holding brake on the drum, changes the picture entirely: the gearbox is now between two driven elements, and the stopping torque passes through the gearing at full load torque rather than reflected. This arrangement is used where a redundant brake is required, but it must be declared at enquiry because the gear train rating changes.

There is also the question of what happens when the brake releases. On a hoist, the load is always trying to run away; the brake must hold before the motor produces torque and release only once it does. That sequencing is an electrical control issue, but it determines whether the gearbox ever sees a reverse torque spike on every lift start. A drive that chatters between brake and motor at each start will show up as tooth wear long before its calculated life is reached.

The Load Path Through the Housing

On a hoist, the output shaft does not only transmit torque. It also carries the rope or chain pull, which acts at a distance from the bearing and produces a bending moment. That moment has to go somewhere: into the bearings, into the housing, and finally into the mounting bolts and the crane structure.

Three consequences follow, and each is a routine source of failure:

The overhung load must be stated, not assumed. A drum mounted directly on the output shaft places the rope pull at a specific distance from the bearing centre; the allowable overhung load falls as that distance grows. Suppliers publish a curve of allowable load against distance, and a unit that is fine at the bearing face can be overloaded at a hundred millimetres out.

The housing must be supported, not just bolted. A gearbox hanging from four bolts on a thin plate will deflect under the bending moment, and the deflection shows up as misalignment between the output shaft and the drum bearing. Rigid mounting pads, or a torque arm where the housing is foot-mounted elsewhere, keep the load path short.

The bearing on the load side usually fails first. It carries the radial load plus whatever axial thrust the rope geometry produces. In maintenance records of hoist drives, the output bearing and the output seal are consistently the earliest wear items, which is why they are the parts worth checking first when a unit starts to run hot.

Efficiency, Backdriving and the Lowering Case

A hoist gearbox has to be considered in both directions.

During lifting, the motor drives through the gearing and efficiency costs power; a two-stage helical unit in the high nineties per stage is typical, a worm unit far lower.

During lowering, the load drives the motor, and the gearing becomes a generator path. If the ratio and friction allow the load to backdrive, the motor must be able to absorb the regenerated energy, usually through a braking resistor or a regenerative drive. If the train is self-locking, the load will not run down on its own, but the drive may then need more torque to lower it, and the brake takes a larger share of the holding duty.

Worm units are sometimes chosen precisely because they self-lock, which is attractive on a simple hoist with no redundant brake. The cost is efficiency, and at higher powers the heat generated inside the housing becomes the limiting factor rather than the tooth stress. Check the thermal rating, not just the mechanical one.

Selection Checklist Before You Send an Enquiry

A drive specified from these eight lines can be quoted and cross-checked in one exchange. A drive specified as "5.5 kW, 12 rpm" cannot, and the resulting quotation is a guess with a delivery date attached.

For hoist and travel drives on industrial cranes, a foot-mounted or flange-mounted helical unit sized on the mechanism group rather than on motor power is the usual starting point; our M3RVSF70 vertical bevel helical gear unit shows how the frame, ratio and mounting options are arranged on a right-angle build, and the same sizing logic applies to inline units. If the duty point is still open, the online selection tool will produce a frame and ratio from torque and speed in a few minutes.

FAQ

What service factor should a hoist gearbox use?
It depends on the mechanism group. General factory hoists in the M4 to M5 range typically use a service factor around 1.5 to 1.8 on nominal torque; M6 and above can require 2.0 or higher. The factor must be applied to the load spectrum, not to the rated capacity alone.

Can I replace a hoist gearbox with a standard industrial unit of the same ratio and power?
Only if the duty class, overhung load, brake arrangement and mounting pattern all match. Ratio and power are the two easiest numbers to check and the two that reveal the least. A unit that matches on paper but has a smaller bearing set or a lighter housing will fail at the output bearing.

Why does my hoist gearbox run hot only when lowering?
Lowering reverses the power flow, and in many drives the gearing runs on the opposite tooth flank with different sliding conditions. Combined with a brake that is doing more holding work, the losses rise. If the housing temperature climbs on lowering but not on lifting, check the brake release timing and the oil level before suspecting the gears.

Does a vertical mounting need a different oil fill?
Yes. An upright housing has its lowest bearing submerged and its upper bearings relying on splash, and the oil level is only meaningful with the unit level and stopped. Filling to the mark while the shaft turns overfills the housing and pushes oil past the top seal.

How often should hoist gearbox oil be changed?
The first change is usually early, after a few hundred running hours, to remove running-in debris. After that, interval depends on duty and oil type, and condition monitoring by sampling is more reliable than a fixed calendar. High-cycle cranes in dusty environments benefit from sampling at each planned inspection.